Skip to content

TST: troubleshoot flaky linux plotting tests#64317

Open
jbrockmendel wants to merge 9 commits intopandas-dev:mainfrom
jbrockmendel:tst-plotting-strict-2
Open

TST: troubleshoot flaky linux plotting tests#64317
jbrockmendel wants to merge 9 commits intopandas-dev:mainfrom
jbrockmendel:tst-plotting-strict-2

Conversation

@jbrockmendel
Copy link
Member

Removed strict=False xfail-on-linux in 4 places to see if they have magically resolved themselves.

Copy link
Member

@mroeschke mroeschke left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I would be fine passing a fairly generous tolerance to the assertion functions just to be safe. AFAICT this keyword is mostly just passed through to Matplotlib?

@jbrockmendel
Copy link
Member Author

Updated to allow a small tolerance.

# GH#64317 on some linux builds this is flaky with the first entry being
# off by -1.69e-21. Rather than xfail this test, we allow a small
# tolerance, as it isn't really user-visible.
tm.assert_almost_equal(result, expected, atol=2e-21)
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Given that the expecteds only is as small as 1e-5, could we bump the atol to like 1e-15 to give us more wiggle room? Just want to rule out floating point artifacts biting us again here

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

sure, will update

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

2 participants